Drug-sensitive FGFR2 mutations in endometrial ... - Semantic Scholar

2 downloads 0 Views 514KB Size Report
Jun 24, 2008 - Med Genet A 140:2245–2247. 34. Rouzier C, et al. (2007) Ovarian dysgerminoma and Apert syndrome. Pediatr Blood. Cancer 50:696 – 698.

Drug-sensitive FGFR2 mutations in endometrial carcinoma Amit Dutt*†, Helga B. Salvesen‡§, Tzu-Hsiu Chen*†, Alex H. Ramos*†, Robert C. Onofrio†, Charlie Hatton†¶, Richard Nicoletti*†, Wendy Winckler†¶, Rupinder Grewal†¶, Megan Hanna†¶, Nicolas Wyhs†¶, Liuda Ziaugra†, Daniel J. Richter†, Jone Trovik‡§, Ingeborg B. Engelsen‡§, Ingunn M. Stefansson储**, Tim Fennell†, Kristian Cibulskis†, Michael C. Zody†, Lars A. Akslen储**, Stacey Gabriel†, Kwok-Kin Wong*††, William R. Sellers‡‡, Matthew Meyerson*†¶§§¶¶, and Heidi Greulich*†,††¶¶ *Department of Medical Oncology and ¶Center for Cancer Genome Discovery, Dana-Farber Cancer Institute, Boston, MA 02115; ††Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA 02115; §§Department of Pathology, Harvard Medical School, Boston, MA 02115; †The Broad Institute of MIT and Harvard, Cambridge, MA 02142; ‡Department of Clinical Medicine and 储The Gade Institute, Section for Pathology, University of Bergen, N-5020 Bergen, Norway; Departments of **Pathology and §Obstetrics and Gynecology, Haukeland University Hospital, N-5020 Bergen, Norway; and ‡‡Novartis Institutes for Biomedical Research, Cambridge, MA 02139

Oncogenic activation of tyrosine kinases is a common mechanism of carcinogenesis and, given the druggable nature of these enzymes, an attractive target for anticancer therapy. Here, we show that somatic mutations of the fibroblast growth factor receptor 2 (FGFR2) tyrosine kinase gene, FGFR2, are present in 12% of endometrial carcinomas, with additional instances found in lung squamous cell carcinoma and cervical carcinoma. These FGFR2 mutations, many of which are identical to mutations associated with congenital craniofacial developmental disorders, are constitutively activated and oncogenic when ectopically expressed in NIH 3T3 cells. Inhibition of FGFR2 kinase activity in endometrial carcinoma cell lines bearing such FGFR2 mutations inhibits transformation and survival, implicating FGFR2 as a novel therapeutic target in endometrial carcinoma. endometrial cancer 兩 fibroblast growth factor receptor 2 兩 oncogene 兩 targeted therapy 兩 tyrosine kinase


yrosine kinases play a major role in transduction of proliferative signals and can become oncogenic when deregulated by somatic mutation (1). Somatically altered tyrosine kinases have proven to be tractable therapeutic targets in several tumor types; examples of successfully targeted tyrosine kinases include ABL1 in chronic myeloid leukemia (2), KIT in gastrointestinal stromal tumors (3), ERBB2 in breast cancer (4), and EGFR in non-small-cell lung cancer (5–7). The tyrosine kinase family has not been exhaustively studied in human cancer, and it is likely that additional tyrosine kinase therapeutic targets remain to be discovered. The fibroblast growth factor receptor (FGFR) tyrosine kinase family, which is comprised of four kinases that differentially respond to 18 FGF ligands (8, 9), has long been implicated in cancer. Translocations involving FGFR3, and activating somatic mutations in FGFR3, have been identified in multiple myeloma patients (10, 11), and translocations of FGFR1 have been found in patients with 8p11 myeloproliferative syndrome (12). Isolated cases of a missense mutation of FGFR4 in a lung adenocarcinoma patient and missense mutations of FGFR2 in a lung squamous cell carcinoma patient and gastric cancer patient have also been reported (13, 14). In addition to these documented examples of somatic mutation of FGFR family members in cancer, a germ-line polymorphism in the second intron of FGFR2 was found to be associated with breast cancer in genomewide association studies (15, 16). FGFR1–FGFR3 are characterized by alternative splicing of the mRNA encoding the third Ig-like repeat in the extracellular ligand-binding domain. This differential splicing determines ligand specificity such that isoforms expressed primarily in epithelial cells (IIIb) preferentially bind FGF ligands expressed www.pnas.org兾cgi兾doi兾10.1073兾pnas.0803379105

by mesenchymal cells, and isoforms expressed primarily in mesenchymal cells (IIIc) preferentially bind FGF ligands expressed by epithelial cells (17–19). Alteration of this restricted expression pattern can lead to oncogenic transformation (20). Mutations in FGFR2 and FGFR3 can also alter ligand specificity (21, 22); such gain-of-function mutations have been found in patients with the congenital craniofacial malformation disorders, Apert and Crouzon syndromes, characterized by craniosynostosis and syndactyly of hands and feet (23, 24). Moreover, these mutations are directly correlated with development of Apert syndrome in mouse models (25–27). In a search for novel druggable therapeutic targets, we have undertaken sequencing of tyrosine kinase genes in a variety of tumor types. Endometrial cancer, the most common gynecological cancer in industrialized countries, with no curative treatment for patients with nonresectable disease (28), was examined in early discovery experiments. Here, we report somatic mutations of FGFR2 in endometrial cancer, with additional instances found in lung squamous cell carcinoma and cervical carcinoma. Results Identification of Cancer-Associated Mutations in Endometrial Carcinoma. We performed DNA sequence analysis of all exons of 89

tyrosine kinase genes and 19 additional known oncogenes and tumor suppressor genes in 40 primary endometrial carcinoma tumor DNA samples. Nonsynonymous candidate mutations detected by sequencing of tumor DNA were confirmed and determined to be somatic or germ-line by mass spectrometric genotyping of the tumor and matched normal DNA. Somatic mutations in genes previously demonstrated to be altered in endometrial carcinoma, including KRAS, CTNNB1, PIK3CA, PTEN, and TP53, were found at the expected frequencies [supporting information (SI) Fig. S1]. One sample was deemed to be hypermutated because of the presence of an unusually large number of somatic mutations in the genes sequenced and was analyzed separately (Table S1). Somatic mutations in 19 additional genes were detected in the nonhypermutated samples, Author contributions: A.D. and H.B.S. contributed equally to this work; A.D., H.B.S., W.R.S., M.M., and H.G. designed research; A.D., H.B.S., T.-H.C., A.H.R., R.C.O., W.W., R.G., N.W., and L.Z. performed research; H.B.S., C.H., R.N., M.H., D.J.R., J.T., I.B.E., I.M.S., T.F., K.C., M.C.Z., L.A.A., S.G. and K.-K.W. contributed new reagents/analytic tools; A.D., H.B.S., T.-H.C., A.H.R., C.H., R.N., M.H., J.T., I.B.E., I.M.S., L.A.A., M.M., and H.G. analyzed data; and A.D., M.M., and H.G. wrote the paper. The authors declare no conflict of interest. ¶¶To

whom correspondence may be addressed. E-mail: matthew㛭[email protected] harvard.edu or [email protected]

This article contains supporting information online at www.pnas.org/cgi/content/full/ 0803379105/DCSupplemental. © 2008 by The National Academy of Sciences of the USA

PNAS 兩 June 24, 2008 兩 vol. 105 兩 no. 25 兩 8713– 8717


Communicated by R. L. Erikson, Harvard University, Cambridge, MA, April 9, 2008 (received for review December 20, 2007)

Fig. 1. Schematic diagram of novel FGFR2 mutations. Black diamonds indicate each instance of mutations found in endometrial samples; red diamonds indicate squamous lung; blue diamonds indicate cervical carcinoma. Asterisks indicate mutations demonstrated to be oncogenic (data in Table S3). K310R and A389T were not transforming; the remaining mutants without asterisks were not tested. Ig, Ig-like repeat. TM, transmembrane domain.

most frequently in the known tumor suppressor gene VHL (Table S2 and Fig. S1). Among the remaining mutated genes, we found that two tumor DNA samples harbored identical mutations in FGFR2, S252W (Table S2 and Fig. S2). Using a combination of sequencing and mass spectrometric genotyping, we identified 13 more samples with FGFR2 mutations among an additional 73 endometrial carcinoma samples and 9 endometrial cancer cell lines (Fig. 1 and Table S3), for a total of 15 samples with FGFR2 mutations of 122 endometrial carcinoma DNAs tested (12.3%). We also found FGFR2 mutations in two of 46 cervical carcinoma DNA samples and two of 42 lung squamous cell carcinoma DNA samples (Fig. 1 and Table S3). No mutations in FGFR2 were detected in other tumor types in which we systematically sequenced tyrosine kinase genes, including lung adenocarcinoma, renal cell carcinoma, glioblastoma, and prostate cancer (data not shown). The somatic FGFR2 mutations include the S252W and P253R alleles, autosomal dominant mutations associated with the congenital developmental disorder Apert syndrome, and N549 and K659 mutations associated with a second congenital developmental disorder, Crouzon syndrome. These latter kinase domain

residues have recently been shown to participate in a ‘‘molecular brake’’ regulating FGFR2 kinase activity (24, 29). Mutant FGFR2 Is Oncogenic. Although Apert syndrome-associated

FGFR2 mutants have been shown to exhibit increased kinase activity, the putative transforming activity of such FGFR2 mutants is only tangentially referred to in the literature (30). We therefore transduced NIH 3T3 fibroblasts with retroviruses encoding isoform IIIb WT FGFR2, FGFR2 S252W, and FGFR2 S252W in combination with the kinase-inactivating D626A substitution. Ectopic expression of the Apert syndromeassociated S252W mutant in pooled NIH 3T3 cells conferred anchorage-independent growth, forming 8-fold more colonies in soft agar than cells expressing WT FGFR2 (Fig. 2A), despite higher expression levels of WT FGFR2 (Fig. 2B). Transformation was accompanied by elevated phosphorylation of the FGFR2 substrate FRS2 (Fig. S3). NIH 3T3 cells expressing FGFR2 S252W in the context of an inactivating D626A mutation did not form colonies in soft agar, indicating that enzymatic activity is required for transformation (Fig. 2 A). Other endometrial cancer-derived FGFR2 mutations were found to be similarly oncogenic (Table S3). Mutant FGFR2 Is Required for Tumor Cell Survival. To determine whether expression of mutant FGFR2 is required for tumor cell survival, we tested a series of shRNA constructs in endometrial tumor cell lines expressing WT FGFR2 (Hec-1B) or mutant FGFR2

Fig. 2. Ectopic expression of FGFR2 S252W in NIH 3T3 cells supports anchorage-independent growth. (A) Pooled NIH 3T3 cells stably expressing the empty vector or the WT, S252W, or kinase-dead S252W/D626A isoform IIIb FGFR2 constructs were assessed for colony formation in soft agar. n ⫽ average number of colonies counted in three sets of 10 fields. (B) FGFR2 expression was confirmed by immunoblotting using actin as a loading control. wt, WT FGFR2. 8714 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0803379105

Fig. 3. Knockdown of mutant FGFR2 expression with shRNA inhibits transformation and survival of endometrial cancer cell lines. Infection with two independent hairpins (shFGFR2#1 and shFGFR2#4) inhibited cell survival as assessed by WST-1 cell survival assay (A and B) and anchorage-independent growth as assessed by colony formation in soft agar (C and D) in the MFE-280 cells harboring an S252W mutation, but not the Hec-1B cells, which express WT FGFR2. shGFP, a hairpin specific for green fluorescent protein, was used as a negative control. All results are normalized to survival or colony formation by cells infected with empty vector.

Dutt et al.

(MFE-280, MFE-296, AN3CA) (Table S3). We identified two hairpin RNAs that efficiently knocked down expression of FGFR2 (Fig. S4A) and used these to demonstrate that knockdown of mutant FGFR2 expression in the MFE-280 cells, but not WT FGFR2 expression in the Hec-1B, inhibited anchorage-independent growth and cell survival (Fig. 3). Similar inhibition of cell survival was observed upon treatment of the MFE-296 and AN3CA cell lines with the FGFR2 shRNA (Fig. S4 B and C). FGFR Inhibitors Block Proliferation and Survival of Cancer Cell Lines Bearing FGFR2 Mutations. We then investigated whether inhibition

of FGFR2 kinase activity would be effective against endometrial carcinoma cell lines bearing the FGFR2 mutations. Treatment of the MFE280, AN3CA, and MFE296 endometrial carcinoma cell lines harboring FGFR2 mutations with the FGFR inhibitor PD173074 (31) resulted in a marked decrease in colony formation in soft agar and cell survival in liquid culture, with IC50s in ranges of 20–30 and 2–40 nM, respectively, whereas similar treatment of the Hec-1B cell line expressing WT FGFR2 had little or no effect (Fig. 4 B–D). Treatment with PD173074 similarly inhibited phosphorylation of FRS2, which was constitutively phosphorylated in the untreated endometrial cell lines harboring activating FGFR2 mutations relative to the Hec-1B (Fig. 4A). Discussion We have identified oncogenic somatic mutations in FGFR2 identical to germ-line alterations in craniofacial malformation syndromes. Activating germ-line mutations in RAS-pathway oncogenes have been shown to be causative in multiple congenital developmental disorders, such as Noonan syndrome, Costello syndrome, and cardio-facio-cutaneous syndrome, each also associated with a predisposition to various malignancies (32). Our results demonstrate a similar link between somatic oncogenic FGFR2 mutations in endometrial, cervical, and squamous lung cancer and identical germ-line FGFR2 mutations in the congenital developmental disorder Apert syndrome and the related Pfeiffer and Crouzon syndromes (24). Surprisingly, only Dutt et al.

isolated cases of malignancies occurring in Apert syndrome patients have been reported (33, 34). FGFR2 mutations have also been reported independently in endometrial cancer (35). The finding of somatic FGFR2 mutations in cancer also provides an interesting convergence with genetic cancer risk: genomewide association studies have implicated germ-line polymorphisms of FGFR2 in susceptibility to breast cancer (15, 16). Our discovery of somatic gain-of-function FGFR2 mutations in three tumor types underscores the general significance of FGFR2 for carcinogenesis. It is noteworthy that FGFR2 has been implicated in cancer both by genomewide association studies and directed sequencing of tumor DNA. The occurrence of somatic FGFR2 mutations in multiple cancers suggests an opportunity for targeted therapy, as the FGFR smallmolecule inhibitor PD173074 diminishes survival and anchorageindependent growth by endometrial cancer cell lines expressing activating FGFR2 mutations. Given additional examples of somatic gain-of-function craniosynostosis syndrome-related FGFR2 mutations such as S267P in gastric cancer (36), systematic sequencing of FGFR2 in a wide variety of tumor types may indicate utility of FGFR2 inhibition in a broad spectrum of cancers. Thus, the FGFR family may join the EGFR family as a widespread target for therapeutic intervention in human cancers. Materials and Methods DNA Samples. Genomic DNA was extracted from surgically dissected and fresh frozen primary tumors from clinically well defined 114 endometrial carcinomas, 46 cervical carcinomas, 42 lung squamous cell carcinoma samples, 9 endometrial tumor cell lines, and 2 lung squamous cell lines. All primary tumor specimens were examined histologically and needle-dissected as needed to ensure at least 80% neoplastic tissue. DNA was isolated by digestion with proteinase K in STE and 10% SDS, followed by a standard phenol-choloroform extraction and ethanol precipitation. Blood or matched nonmalignant tissue was available for all 203 patients, and genomic DNA was extracted as above. Matching of the germ-line and tumor DNA for patient sample pairs was confirmed by mass spectrometric genotyping of 24 SNP loci (see genotyping methods below). Collection and analysis of endometrial and cervical clinical samples was approved by the Norwegian Data Inspectorate, Norwegian Social Sciences Data Services, and the Norwegian Local Ethical Committee. PNAS 兩 June 24, 2008 兩 vol. 105 兩 no. 25 兩 8715


Fig. 4. Inhibition of FGFR2 kinase activity inhibits FRS2 phosphorylation and transformation and survival of endometrial cancer cell lines. (A) FGFR2 substrate FRS2 is constitutively phosphorylated in the MFE-280, AN3CA, and MFE-296 endometrial cancer cell lines harboring FGFR2 mutations, as compared with the Hec-1B line, which expresses WT FGFR2. (Upper) Treatment of these cell lines for 40 min with 2 ␮M FGFR inhibitor PD173074 inhibits both basal and ligand-induced (20-min stimulation with 30 ng/ml FGF7) phosphorylation, as evidenced by immunblotting with anti-phospho-FRS2. (Lower) Similar levels of expression of FRS2 were confirmed by immunoblotting with anti-FRS2. (B) Treatment with the indicated concentrations of PD173074 inhibited soft agar colony formation by the MFE-280, AN3CA, and MFE-296 endometrial cancer cell lines harboring FGFR2 mutations, as compared with the Hec-1B line, which expresses WT FGFR2. Colonies were photographed after 2 weeks. (Magnification: ⫻2.) (C) Quantification of effects of PD173074 on soft agar colony formation with EC50s indicated. (D) Endometrial cancer cell line WST survival assays performed after 4 days of treatment with PD173074. EC50s are indicated.

Whole-Genome Amplification. Primary tumors and matched normals were whole-genome-amplified for sequencing and genotyping. Genotype validation of candidate mutations was performed on an independent whole-genome amplification. Forty nanograms of genomic DNA was used for whole-genome amplification using the REPLI-gTM kit (Qiagen), yielding ⬇40 ␮g of total DNA. Sequencing. PCR primers were designed with which to amplify each exon of interest (Table S4). PCRs for each exon and flanking intronic sequences contained 5 ng of whole-genome-amplified DNA, 1⫻ HotStar buffer, 0.8 mM dNTPs, 2.5 mM MgCl2, 0.2 units of HotStar Enzyme (Qiagen), and 0.25 ␮M forward and reverse primers in a 6- or 10-␮l reaction volume. PCR cycling parameters were: one cycle of 95°C for 15 min; 35 cycles of 95°C for 20 s, 60°C for 30 s, and 72°C for 1 min; followed by one cycle of 72°C for 3 min. The resulting PCR products were sequenced by using bidirectional dye-terminator fluorescent sequencing with universal M13 primers. Sequencing fragments were detected via capillary electrophoresis with an ABI Prism 3730 DNA Analyzer (Applied Biosystems). Sequencing traces were analyzed by using an automated pipeline consisting of SNPCompare (described below) and Mutation Surveyor 2.51 (SoftGenetics), followed by manual review of candidate mutations. Base coverage was determined by using SNPCompare. An exon was considered covered if 80% of the bases were covered in 80% of the samples. Minimum exon coverage was 80%. PCR and sequencing of MFE-280 and Hec-1B were additionally performed at Agencourt Bioscience, Beverly, MA. SNPCompare. SNPs were called by using PolyPhred 6.02b (37, 38) followed by a postprocessing pipeline we refer to as SNPCompare. In brief, we retain all SNPs called by PolyPhred with a confidence of 99. SNPs with a confidence ⱖ95 are kept if the average local read quality is at least 30, whereas those with confidence ⱖ90 are kept if average local read quality is at least 40. For SNPs with PolyPhred confidence ⱖ60 but not meeting the preceding criteria, we check traces with an independent SNP detection method, PolyDhan (D. Richter, unpublished data; see below). PolyDhan has relatively poor sensitivity but very low false discovery, so we retain low confidence PolyPhred SNPs also seen with PolyDhan in the same sample marked heterozygous by PolyPhred. We define local read quality as the average Phred (39, 40) quality score in a ‘‘notched’’ window of length 11 centered on the putative SNP but excluding the SNP base itself and the immediate flanking base on either side (whose quality is expected to be low if the base is a true heterozygote). Average local read quality at a site is computed by averaging the local read quality of all reads aligning to reference. If any sample at an alignment position is called nonreference according to the above criteria, we consider the site to be variant and accept PolyPhred’s genotype for all samples at that site without further quality assessment. Briefly, PolyDhan compares the processed trace signal for all reads on a given strand at a given site in an alignment and determines whether some subset of traces show evidence of a nonreference signal in excess of the average level of background signal in that channel at that base. Both strands are examined independently, and a SNP is called if either strand shows evidence. As used in SNPCompare, PolyDhan is required to confirm one or more heterozygous calls by PolyPhred only at sites not meeting other quality filtering criteria. We do not require PolyDhan to confirm every heterozygous call once the site is accepted as variant. Mass Spectrometric Genotyping. Multiplex PCR was performed in 5-␮l volumes containing 0.1 units of Taq polymerase, 5 ng of whole-genome-amplified genomic DNA, 2.5 pmol of each PCR primer, and 2.5 ␮mol of dNTP. Thermocycling was at 95°C for 15 min followed by 45 cycles of 95°C for 20 s, 56°C for 30s, 72°C for 30 s. Unincorporated dNTPs were deactivated by using 0.3 unit of shrimp alkaline phosphatase followed by primer extension using 5.4 pmol of each primer extension probe, 50 ␮mol of the appropriate dNTP/ddNTP combination, and 0.5 unit of Thermosequenase. Reactions were cycled at 94°C for 2 min, followed by 40 cycles of 94°C for 5 s, 50°C for 5 s, 72°C for 5 s. After addition of a cation exchange resin to remove residual salt from the reactions, 7 nl of the purified primer extension reaction was loaded onto a matrix pad (3-hydroxypicoloinic acid) of a SpectroCHIP (Sequenom). SpectroCHIPs were analyzed with a Bruker Biflex III MALDI-TOF mass spectrometer (SpectroREADER; Sequenom). Cell Culture and Reagents. NIH 3T3 cells obtained from ATCC were maintained in DMEM (Cellgro/Mediatech) supplemented with 10% calf serum (Gibco/ Invitrogen) and penicillin/streptomycin (Gibco/Invitrogen). MFE-280 and Hec-1B cells were maintained in 40% RPMI medium 1640 ⫹ 40% MEM ⫹ 20% FBS and DMEM/F12 at 1:1 ⫹ 15% FBS, respectively. Unless otherwise noted, cells were placed in media containing 0.5% calf serum 24 h before 30 ng/ml FGF7 (Peprotech) stimulation for 20 min at 37°C. PD173074 was purchased from Calbiochem and diluted in DMSO to the indicated concentrations. 8716 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0803379105

Expression Constructs. FGFR2b cDNA clone (IOH36372) was obtained from Invitrogen. Point mutations were made by using the QuikChange Mutagenesis XL kit (Stratagene) with the following mutation-specific oligonucleotide primers: S252W, 5⬘- GACGTACCTTATCTATAAGAGGACCGAC-3⬘ and 5⬘- AGTATTCGTGGTCGGTCCTCTTATA-3⬘; K310R, 5⬘-TGCCCTACCTCAGGGTTCTCAAGGC-3⬘ and 5⬘-GCCTTGAGAACCCTGAGGTAGGGCA-3⬘; A314D, 5⬘GGTTCTCAAGGACGCCGGTGTTAAC-3⬘ and 5⬘-GTTAACACCGGCGTCCTTGAGAACC-3⬘; A389T, 5⬘-TCTTAATCGCCTATATGGTGGTAACAG-3⬘ and 5⬘CTGTTACCACCATATAGGCGATTAAGA-3⬘; N549K, 5⬘-CATAAATCTTCTGGGAGCCTGCACAC-3⬘ and 5⬘-GTG TGCAGGCTCCCAGAAGATTTATG-3⬘; C382R, 5⬘-GATAGCCATTTACCGCATAGGGGTC-3⬘ and 5⬘-GACCCCTATGCGGTAAATGGCTATC-3⬘; D101Y, 5⬘-CGC CACGCCTAGATACTCCGGCCTC-3⬘ and 5⬘-GAGGCCGGAGTATCTAGGCGTGGCG-3⬘; P253R, 5⬘-TGGAGCGATCGCGTCACCGGCCCATC-3⬘ and 5⬘-GATGGG CCGGTGACGCGATCGCTCCA-3⬘. Transfection and Infection. Replication-incompetent retroviruses were produced from pBabe-Puro-based vectors by transfection into the Phoenix 293T packaging cell line (Orbigen) by using Lipofectamine 2000 (Invitrogen). Cells were infected with these retroviruses in the presence of polybrene. Two days after infection, 2 ␮g/ml puromycin (Sigma) was added, and pooled stable cell lines were selected, from which clonal cell lines were derived. shRNA-Mediated FGFR2 Knockdown. shRNA vectors targeted against FGFR2 and GFP were obtained from The RNAi Consortium. The target sequences of the FGFR2 shRNA constructs are: FGFR2#1 (TRCN0000000367), 5⬘-GCCACCAACCAAATACCAAATCTC-3⬘; FGFR2#2 (TRCN0000000368), 5⬘-CCGAATGAAGAACACGACCAA-3⬘; FGFR2#3 (TRCN0000000369), 5⬘-CCCAACAATAGGACAGTGCTT-3⬘; and FGFR2#4 (TRCN0000000370), 5⬘-GCCAACCTCTCGAACAGTATTC-3⬘ The sequence targeted by the GFP shRNA is 5⬘-GCAAGCTGACCCTGAAGTTCAT-3⬘. Lentiviruses were made by transfection of 293T packaging cells with these constructs by using a three-plasmid system as described (41). Target cells were incubated with lentiviruses for 6 h in the presence of 8 ␮g/ml polybrene and left in their fresh respective media. Two days after infection, puromycin (2 ␮g/ml for MFE-280, Hec-1B, and AN3CA and 3.0 ␮g/ml for MFE-296) was added. Cells were grown in the presence of puromycin for 4 days. Fifty micrograms of total cell lysates prepared from the puro-selected cell lines was analyzed by Western blotting using anti-FGFR2 mAb (Santa Cruz Biotechnology) for MFE-296, AN3CA, and Hec-1B and anti-FGFR2 (S252W) polyclonal antibody for MFE-280 tumor cell line, obtained from Bethyl Laboratories and anti-actin mAb (sc-1615; Santa Cruz Biotechnology). Cell Survival Assays with Tumor Cell Lines Expressing shFGFR2 and shGFP Constructs. One thousand cells for each tumor cell line expressing shRNAs targeting FGFR2 or GFP along with uninfected cells were seeded in six wells on a 96-well plate. Cell viability was determined at 24-h time points for 4 consecutive days by using the WST-1 assay (Roche Applied Science). The percentage of cell viability was plotted for each cell line of readings obtained on day 3 relative to day 1. Soft Agar Anchorage-Independent Growth Assay. FGFR2-expressing NIH 3T3 cells were suspended in a top layer of DMEM containing 10% calf serum and 0.4% Select agar (Gibco/Invitrogen) and plated on a bottom layer of DMEM containing 10% calf serum and 0.5% Select agar. PD173074 was added as described to the top agar. After 3 weeks of incubation, NIH 3T3 colonies were counted in triplicate wells from 10 fields photographed with a ⫻4 objective. Endometrial cancer cell lines (with and without shRNA constructs) MFE-280, MFE-296, Hec1B, and AN3CA were suspended in a top layer of media described above with 0.4% Select Agar and otherwise assayed as for NIH 3T3 cells, but photographed with a ⫻2 objective after 2 weeks of incubation. EC50s were determined by nonlinear regression with Prism GraphPad software. Cytotoxicity Assays. Endometrial cancer cell lines were treated with PD173074 1 day after plating, and cell survival was assessed 4 days later by using the WST-1 assay (Roche). Each data point represents the median of six replicate wells for each tumor cell line and PD173074 concentration. EC50s were determined by nonlinear regression using Prism GraphPad software. Immunoblotting. Cells were lysed in a buffer containing 50 mM Tris䡠HCl (pH 7.4), 150 mM NaCl, 2.5 mM EDTA, 1% Triton X-100, and 0.25% Ipegal. Protease inhibitors (Roche) and phosphatase inhibitors (Calbiochem) were added before use. Samples were normalized for total protein content. Lysates were boiled in sample buffer, separated by SDS/PAGE on 8% polyacrylamide gels, transferred to PVDF membrane, and probed as described. Antibodies used for immunoblotting were: anti-FGFR2 (sc-6930; Santa Cruz Biotechnology), anti-phospho-FRS2 Y436

Dutt et al.

ACKNOWLEDGMENTS. We thank Gerd Lillian Hallseth, Bendik Nordanger, and Britt Edvardsen for excellent technical assistance; Jordi Barretina for helpful discussions; and Derek Chiang for critical reading of the

manuscript. This work was supported by Swiss National Science Foundation Fellowship PBZHB-106297 (to A.D.), National Cancer Institute Cancer Genome Characterization Center Grant U24 CA126546 (to M.M.), Helse Vest Grant 911351 (to H.B.S.), the University of Bergen Harald Andersen’s Fund, the Norwegian Cancer Society, the Novartis Foundation, and Novartis Pharmaceuticals.

1. Blume-Jensen P, Hunter T (2001) Oncogenic kinase signaling. Nature 411:355–365. 2. Druker BJ, et al. (2001) Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 344:1031–1037. 3. Demetri GD, et al. (2002) Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 347:472– 480. 4. Slamon DJ, et al. (2001) Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 344:783–792. 5. Paez JG, et al. (2004) EGFR mutations in lung cancer: Correlation with clinical response to gefitinib therapy. Science 304:1497–1500. 6. Lynch TJ, et al. (2004) Activating mutations in the epidermal growth factor receptor underlying responsiveness of nonsmall-cell lung cancer to gefitinib. N Engl J Med 350:2129 –2139. 7. Pao W, et al. (2004) EGF receptor gene mutations are common in lung cancers from ‘‘never smokers’’ and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci USA 101:13306 –13311. 8. Ornitz DM, et al. (1996) Receptor specificity of the fibroblast growth factor family. J Biol Chem 271:15292–15297. 9. Zhang X, et al. (2006) Receptor specificity of the fibroblast growth factor family: The complete mammalian FGF family. J Biol Chem 281:15694 –15700. 10. Chesi M, et al. (1997) Frequent translocation t(4;14)(p16.3;q32.3) in multiple myeloma is associated with increased expression and activating mutations of fibroblast growth factor receptor 3. Nat Genet 16:260 –264. 11. Richelda R, et al. (1997) A novel chromosomal translocation t(4;14)(p16.3;q32) in multiple myeloma involves the fibroblast growth factor receptor 3 gene. Blood 90:4062– 4070. 12. Reiter A, et al. (1998) Consistent fusion of ZNF198 to the fibroblast growth factor receptor-1 in the t(8;13)(p11;q12) myeloproliferative syndrome. Blood 92:1735–1742. 13. Marks JL, et al. (2007) Mutational analysis of EGFR and related signaling pathway genes in lung adenocarcinomas identifies a novel somatic kinase domain mutation in FGFR4. PLoS ONE 2:e426. 14. Davies H, et al. (2005) Somatic mutations of the protein kinase gene family in human lung cancer. Cancer Res 65:7591–7595. 15. Easton DF, et al. (2007) Genomewide association study identifies novel breast cancer susceptibility loci. Nature 447:1087–1093. 16. Hunter DJ, et al. (2007) A genomewide association study identifies alleles in FGFR2 associated with risk of sporadic postmenopausal breast cancer. Nat Genet 39:870 – 874. 17. Miki T, et al. (1992) Determination of ligand-binding specificity by alternative splicing: Two distinct growth factor receptors encoded by a single gene. Proc Natl Acad Sci USA 89:246 –250. 18. Yan G, Fukabori Y, McBride G, Nikolaropolous S, McKeehan WL (1993) Exon switching and activation of stromal and embryonic fibroblast growth factor (FGF)-FGF receptor genes in prostate epithelial cells accompany stromal independence and malignancy. Mol Cell Biol 13:4513– 4522. 19. Ornitz DM, Itoh N (2001) Fibroblast growth factors. Genome Biol 2:REVIEWS3005. 20. Miki T, et al. (1991) Expression cDNA cloning of the KGF receptor by creation of a transforming autocrine loop. Science 251:72–75. 21. Anderson J, Burns HD, Enriquez-Harris P, Wilkie AO, Heath JK (1998) Apert syndrome mutations in fibroblast growth factor receptor 2 exhibit increased affinity for FGF ligand. Hum Mol Genet 7:1475–1483.

22. Yu K, Herr AB, Waksman G, Ornitz DM (2000) Loss of fibroblast growth factor receptor 2 ligand-binding specificity in Apert syndrome. Proc Natl Acad Sci USA 97:14536 – 14541. 23. Ibrahimi OA, Chiu ES, McCarthy JG, Mohammadi M (2005) Understanding the molecular basis of Apert syndrome. Plast Reconstr Surg 115:264 –270. 24. Wilkie AO (2005) Bad bones, absent smell, selfish testes: The pleiotropic consequences of human FGF receptor mutations. Cytokine Growth Factor Rev 16:187–203. 25. Chen L, Li D, Li C, Engel A, Deng CX (2003) A Ser252Trp [corrected] substitution in mouse fibroblast growth factor receptor 2 (Fgfr2) results in craniosynostosis. Bone 33:169 – 178. 26. Wang Y, et al. (2005) Abnormalities in cartilage and bone development in the Apert syndrome FGFR2(⫹/S252W) mouse. Development 132:3537–3548. 27. Hajihosseini MK, Wilson S, De Moerlooze L, Dickson C (2001) A splicing switch and gain-of-function mutation in FgfR2-IIIc hemizygotes causes Apert/Pfeiffer syndromelike phenotypes. Proc Natl Acad Sci USA 98:3855–3860. 28. Pectasides D, Pectasides E, Economopoulos T (2007) Systemic therapy in metastatic or recurrent endometrial cancer. Cancer Treat Rev 33:177–190. 29. Chen H, et al. (2007) A molecular brake in the kinase hinge region regulates the activity of receptor tyrosine kinases. Mol Cell 27:717–730. 30. Mansukhani A, Bellosta P, Sahni M, Basilico C (2000) Signaling by fibroblast growth factors (FGF) and fibroblast growth factor receptor 2 (FGFR2)-activating mutations blocks mineralization and induces apoptosis in osteoblasts. J Cell Biol 149:1297–1308. 31. Mohammadi M, et al. (1998) Crystal structure of an angiogenesis inhibitor bound to the FGF receptor tyrosine kinase domain. EMBO J 17:5896 –5904. 32. Schubbert S, Shannon K, Bollag G (2007) Hyperactive Ras in developmental disorders and cancer. Nat Rev Cancer 7:295–308. 33. Andreou A, et al. (2006) Early onset low-grade papillary carcinoma of the bladder associated with Apert syndrome and a germline FGFR2 mutation (Pro253Arg). Am J Med Genet A 140:2245–2247. 34. Rouzier C, et al. (2007) Ovarian dysgerminoma and Apert syndrome. Pediatr Blood Cancer 50:696 – 698. 35. Pollock PM, et al. (2007) Frequent activating FGFR2 mutations in endometrial carcinomas parallel germ-line mutations associated with craniosynostosis and skeletal dysplasia syndromes. Oncogene 26:7158 –7162. 36. Jang JH, Shin KH, Park JG (2001) Mutations in fibroblast growth factor receptor 2 and fibroblast growth factor receptor 3 genes associated with human gastric and colorectal cancers. Cancer Res 61:3541–3543. 37. Bhangale TR, Stephens M, Nickerson DA (2006) Automating resequencing-based detection of insertion-deletion polymorphisms. Nat Genet 38:1457–1462. 38. Stephens M, Sloan JS, Robertson PD, Scheet P, Nickerson DA (2006) Automating sequence-based detection and genotyping of SNPs from diploid samples. Nat Genet 38:375–381. 39. Ewing B, Green P (1998) Base-calling of automated sequencer traces using Phred. II. Error probabilities. Genome Res 8:186 –194. 40. Ewing B, Hillier L, Wendl MC, Green P (1998) Base-calling of automated sequencer traces using Phred. I. Accuracy assessment. Genome Res 8:175–185. 41. Moffat J, et al. (2006) A lentiviral RNAi library for human and mouse genes applied to an arrayed viral high-content screen. Cell 124:1283–1298.

Dutt et al.

PNAS 兩 June 24, 2008 兩 vol. 105 兩 no. 25 兩 8717


(3861; Cell Signaling Technologies), anti-phospho-FRS2 Y196 (3864; Cell Signaling Technologies), and anti-FRS2 (sc-17841; Santa Cruz Biotechnology).

Suggest Documents